BIO 432 Taxonomy or Systematics?

BIO 432

Field Botany

Phil

Ganter

301 Harned Hall

963-5782

Restinga Jurubatiba in Brazil The restinga biome (a tropical dune forest) is both valuable (it contains a diversity of valuable plants and animals) and disappearing. The restinga is part of the Atlantic Coastal Forest biome in Brazil, one of the most diverse terrestrial plant communities. Less than 10% of the original area occupied by this biome is undisturbed by man

Taxonomy is the science of naming organisms based on accepted criteria (the criteria are derived from the scheme underlying the taxonomy). This sounds pretty cut and dry but, as in many things, the devil is in the details. As a biologist, you should be interested in the science of taxonomy, even if you have never given it serious consideration or have done so and dismissed it as uninteresting (trivial? prehistoric? fussy?). You certainly make use of taxonomy. We cannot make a systematic study of things (living or otherwise) without a system for organizing our knowledge. There are just too many things. Taxonomy happens whenever the number of things gets too large to allow them to be discussed clearly. It happened to geneticists. A consistent naming scheme for loci and alleles was adopted after the number of genes identified grew too large. It even happened to atomic physicists after the atom smashers started to find heaps of subatomic particles. (By the way, an eminent particle physicist, questioned about all of the new particles, admitted to being flustered by their proliferation. He is reported to have replied that, if he were able to keep track of so many things, he would have become a botanist). Certainly, any biologist working in natural systems is in great need of some taxonomy. Anyone can see that there are lots of different plants, even in simple communities. The question considered here is "how should we go about organizing groups of living things?"

One way of organizing knowledge about differences between plants is to name the plants in such a way that their names convey some additional information beyond the name. Certainly, taxonomists are not satisfied to merely label organisms. This point may not be so obvious. Suppose you have a friend who has just acquired two dogs that he named Bob and Larry. One is large and aggressive and the other small and gentle. When you see the two dogs, could you tell which is Bob and which is Larry? No. In this case, the names convey no information and merely serve as tags to identify the dogs. Suppose your friend had named the dogs Killer and Daisy. When you saw the same two dogs with these names, you could make an informed guess as to which was Killer and which Daisy (assuming you know your friend well enough to know if he or she is a wise guy or not). So, names can be more than labels. They can convey additional information. Taxonomy gives names to organisms that convey information about the organism. If I tell you that the cactus in the picture below is Pilosocereus royenii, I have given you a wealth of information about the organism in the picture.

I have told you something about its external features (all Pilosocereus have similar flowers, similar overall growth form, and similar tufts of hairs along the ribs), its internal features (all have similar secondary chemistry and anatomic details), and where it comes from (P. royenii is only found on some Caribbean islands). Of course, you have to know something about the details of naming cacti to know all of this, but once known, names almost become stories about organisms, not just names. Recently, a friend sent me a picture of someone that had cacti in the background. From the cacti, I could tell him where the picture was taken (the tip of the Baja California peninsula). Taxonomy is constantly featured in episodes of the popular TV series CSI for this reason.

Taxonomy and evolutionary biology have become linked, although each has an independent origin and history. As evolution became accepted by the the majority of biologists, taxonomy and evolution became fused in the field of systematics. Systematics is the science of naming and organizing organisms based on common ancestry (= evolutionary relatedness). In a systematic taxonomy, all of the species in a genus are more related (= have a more recent ancestor in common) to one another than to any other species. Most taxonomies used today are systematic taxonomies, although some persist that are not. (If you think of the reptiles as a taxonomic group, then non-systematic taxonomy persists in you!) An example of scientists using non-systematic taxonomy comes from the standards for describing fungi (as published in the Botanical Code). When describing a new yeast, if I can not find any sexual characteristics, then I must place the species into a genus with other asexual species, even if gene sequence data indicate that the new species is very closely related to a sexual species. The asexual genus is a hodgepodge of unrelated species. Thus, much of fungus taxonomy is not systematic. When describing a sexual yeast, a systematic approach is allowed. This inconsistency is the product of history. Until genetic data became widely available, it was not clear to which genus many asexual species belonged. I predict that all fungal taxonomy will be systematic in the near future, but it hasn't happened as of yet.

Often, systematic taxonomies are referred to as "natural classifications." This means that the taxonomy reflects groups that exist in nature. The taxonomist is just describing groups that already occur. If you have a group of cats to classify, placing all lions in one group, all house cats in a group, and all tigers in another is a natural classification. However, perhaps you are classifying the cats at a zoo and you are in charge of feeding them. In this instance, placing all cats under 30 pounds in one group and all cats over 30 pounds in another is a useful taxonomy but it is not seen as natural. Non-natural classifications are the result of decisions made by the taxonomist but they do not reflect the natural situation.

Usually, you see a scientific name reported as a binomial (which simply means a two-part name), like Syringia vulgaris (the garden lilac). Latin and Greek are used (and a good binomial doesn't mix the two). When a term is not Latin or Greek it is Latinized. (Opunita engelmannii was named for a botanist named Engelmann). This makes the scientific name a foreign term in English and so scientific names are italicized (as are all foreign terms in English). However, the binomial is really only the final two parts of a series of names, as the first part is the name of the genus to which the organism belongs and the second is its species. Perhaps we use the binomial because we often use two names for people and using two for an organism then seems to identify the organism as though it were a person. This is a bit misleading. When we name a person John Doe, we are attempting to give an individual a name that belongs to no one else. When we call an oak tree Quercus emoryi, we mean it belongs to the genus and species named. There are many Quercus emoryi trees and a different scheme is needed to identify individual plants.

Traditionally, taxonomists developed a hierarchical scheme for naming organisms. A hierarchy is an organizational scheme based on a series of levels. Each level has one or more groups (by definition, the highest level has only one group). At any level, each group is composed of groups from the next level down. Thus, a genus like Quercus (the oaks) is composed of groups from the next level down (species in this case). All of this might seem so obvious as to be trivial. However, hierarchies are not the only way to organize things (although they seem to be the scheme of choice for many areas of human knowledge). We will not spend time here on other ways to organize but biologists have made use of them. Behavioral biologists have discovered that some animal societies are strictly hierarchical and that others have little hierarchy, although they do not lack organization. Ecologists have proposed that some communities of competing species might be more stable if they are not organized hierarchically.

There is more to traditional taxonomy than its hierarchical scheme. How groups are formed is key to understanding the plan and understanding the plan is key to getting the information available from a scientific name. Taxonomic groups are formed on the basis of members being similar to one another. This similarity is not a vaguely defined thing but is based on particular characteristics which all members of a group must share. Such characteristics are called key characteristics. Thus there are specific key characteristics that all Quercus emoryi have and generic key characteristics all Quercus have. A northern red oak (Quercus rubra) shares generic key characteristics with an Emory oak (Quercus emoryi) but not specific key characteristics (note that I use the adjective "specific" to mean "having to do with biological species" and not the more usual "having to do with this particular thing"). Ideally, key characteristics should be unique to that group. This is sometimes difficult to do but in any case, key characteristics must belong only to the members of a one group (say one species) within a larger group (the genus to which that species belongs).

Descriptions of species in books and on the web often have many characteristics listed that are not key characteristics. My tree identification book says northern red oaks have leaves 5 to 8 inches long. It also says swamp chestnut oaks have leaves with the same range of lengths, so leaf length is not a key characteristic for either species of oak. Key characteristics can be found along with a host of other characteristics of the group in the scientific article that first defined the group or the most recent article that re-defined the group. (When we learn more about the organisms, redefinition is often necessary). They are often difficult to use unless you are in a laboratory as they can be microanatomic or chemical characteristics. When identifying a plant, you will usually use a series of non-key characteristics to decide on membership in a species. So, if you have four different oak species, one may have 5 inch leaves and smooth acorn cups, another may have 5 inch leaves and rough cups, a third may have 6 inch leaves and rough cups and a fourth 6 inch leaves and smooth cups. Neither leaf size or acorn cup texture is a key characteristic because no group has a unique leaf size or cup texture. Even so, these characteristics can be used in combination to separate the four species.

In addition to defining key characteristics, taxonomists try to maintain consistency when describing new taxa, whether they are families, species or genera. By consistency, I mean that they prefer to organize species into genera or individuals into species such that the level of similarity between species within each genus or individuals within each species is the same. Imagine you are describing eight different species that do not belong to any known genus. But how many new genera to form? You feel that the most unlike of the eight species are too dissimilar to belong to the same genus so more than one new genus will be necessary. Upon inspection, you find that there are two clusters of species (one with three species and the other with four) that are distinct from one another but with little difference between species within each cluster. These seven species will form two new genera but what to do with the eighth, which is not very similar to the other seven species? Usually, it will become a "monospecific" genus rather than just being included in with the genus to which it is most similar. Consistency can be hard to maintain. Suppose you discover an ninth species. It is most similar to the eighth species that is now in its own genus, but the level of similarity between the new species and the eighth is not as high as the level of similarity within the other two genera. If you include the new species in the same genus with the eighth, consistency has been undermined because now one genus has somewhat more dissimilar species in it than the others. If you describe a fourth genus, the number of genera, etc. will seem to grow exponentially fast.

Consistency is very hard to maintain when the number of species in a group gets large or when describing different types of organisms. The morphological differences between the species of the genus Homo (to which you belong) is large when compared to the differences between species within frog genera. Consider the difference in size for chimpanzees and gorillas or the skull shape of gorillas and man. Frog species with such large morphological differences as there are between species of Homo are usually put into different genera or even different families. The rules for separating frogs into different species are different from those for separating primates into species, which is not consistent. A second problem for the goal of consistency is that different kinds of characteristics may support the formation of different taxa. The morphological difference between man and chimpanzee is great but the genetic distance (measured as the average difference between the basepair sequence for a sample of genes from each species) is small. Many vertebrate species include individuals that are genetically more dissimilar than the average man and chimp. The growing use of gene sequence data often supports previously defined taxa but there are important cases in which the new data is not consistent with older data. What to do then?

The problem goes even deeper than what to do when different types of data disagree. Notice that in the explanation of what taxonomists do above, there is a lot of decision making by the taxonomists. The researcher must decide if all characteristics are equally important or if some characteristics carry more relevant information. When examined in detail, it becomes obvious that decisions must be made at each step of the process of constructing a taxonomic scheme. This has long bothered many who study organisms above the level of the individual. Given the same set of data, two taxonomists might easily decide to group the organisms differently. Different conclusions from the same data is problematic. There have been two general responses to this problem: phenetics and cladistics. The first is not a systematic approach (that it, based on relatedness), the second one is.

Pheneticists approach taxonomy by measuring as many characteristics of each individual as possible. They collect large data sets in which there are bound to be inconsistencies. They then apply a mathematical algorithm designed to form groups from this large mass of data. The math models have become both powerful and complex as the field has developed. They have the useful property that, given the same data, two different pheneticists will form the same groups if both use the same algorithm to do the grouping. Consistency is still not achieved because there are lots of models, which means that the pheneticist must make a decision about which model to use and this is a bit more decision making than purists will accept. In any case, these models are not based on any ideas about ancestry of the organisms studied, so the taxonomy produced is not systematic. In a phenetic system, one might measure flower color as one characteristic. It is useful for separating flowering plants into groups. This may not be enough to qualify flower color as a useful systematic characteristic. It is useful in systematics only if those flowers with the same color are similar because they are descended from an ancestor with that color flower. Thus, flower color is determined by ancestry (by the way, flower color is sometimes useful and sometimes not to systemetists).

Cladistics is a particular systematic scheme based on an explicit theory of evolution. In cladistics., organisms inherit their genes through meiosis or mitosis (so it applies to sexual and asexual organisms). This is the basis for the focus on ancestry as linking organisms into groups. As change occurs, a group of organisms splits into two different branches (= clades, hence the odd name for this method). Thus, a group grows like a tree, with the ancestral group being the trunk. As you go from the base of the tree upward, you are traveling through time towards the present. The trunk splits (= bifurcates because the splits are always into two new branches) and the original group is now replaced with two new groups, each of which may split. So the tree grows more and more branches as time passes. The tips of the last splits (the tiniest twigs on the tree) represent the most recent clades, the only clades still in existence. The tree metaphor is a very good one. In fact, the result of a cladistic analysis is a tree-like diagram called a dendrogram (the root term "dendro" comes from the Greek dendron for tree).

Cladistics is a set of methods to uncover the tree, which is all in the past (by definition), using characteristics measured from the organisms alive today. Since the methods are standardized, the same data will produce the same groupings. There are a lot of different methodologies, but the problem of multiple methods is not as difficult for cladists as it is for pheneticists. This is because systematics is a natural system, so the different methods are all uncovering the same natural groups. Thus, the assumption is that they should all produce the same results if the data and methods are free of error. To construct a cladogram, one must measure characteristics that indicate common ancestry. Whale flippers, primate hands, and frog paws all have five fingers. The development of these organs supports the hypothesis that they are similar because the ancestor of these organisms had five fingers. You may recognize this similarity-due-to-descent as homology. Lets go back to flower color. Flowers in cool tropical uplands are more often red than expected. Is this because the plants in this habitat are all related? In this case, probably not. Lots of very different plants occur in these areas. However, cool areas like these have fewer flying insects and many pollinators are not insects but vertebrates. Reds are apparently easier for vertebrates to see so the increased incidence of red flowers is probably related to the utility of having a flower that is easily seen. In this case, red flower color does not necessarily indicate common ancestry.

Deciding which characteristics contain information about ancestry can be tricky. This, and the large number of calculations needed to analyze data, limited cladistic analyses for many years. However, two technological advances have turned it into a preferred means of making natural classifications. First, computers have made it possible to do the many trillions of calculations often needed to perform an analysis. Second, DNA sequencing technology has given researchers the means to measure characteristics (gene sequences) that are the very stuff of common descent. The details can get a bit tricky but, in general, the assumption that two individuals have the same gene sequence because they share a common ancestor is a reasonable one.

How, you might ask, does all of this affect you in this class? In articles, book and websites devoted to plant classification, you will see that it either uses a standard taxonomic or a cladistic classification. You have to recognize which is being used and to appreciate the reasons why the two schemes are not equivalents. If you go to the USDA site to classify your plants, you will get a standard taxonomic approach, where there are species, genera, families, etc. "The Tree of Life" site (notice the use of the metaphorical tree again) is a second place to get an overall classification of plants that uses a cladistic scheme. They are not the same. Why? The USDA site is based on a taxonomic scheme that tries to fit plants into a preset hierarchy. Those who have constructed this hierarchy (many many researchers working over many many years) have tried to use characteristics that reflect common ancestry (so this is an attempt at a natural system) but made groups according to a pre-ordained taxonomic structure. This approach is called evolutionary systematics (a bit of a redundancy). Thus, all plants in the family Fabaceae (pea family) are believed to share a common ancestor, i. e. to be more related to one another than to other plants. The "Tree of Life" site presents a strictly cladistic analysis, in which the branching pattern arises from the data, not from a preset taxonomic hierarchy. A cladogram is not the same as the standard taxonomic hierarchy and can be very different from it. Just try it for the plants. The USDA site presents the plants like an outline, in which each lower taxonomic level is indented from the next higher level. Go there and see at http://plants.usda.gov/cgi_bin/topics.cgi?earl=classification.html. Select "Class" as the rank and you will get a hierarchical classification of all plants to the level of the family. I have reproduced the results below.

If you go the the Tree of Life web page on embryophytes, the nearest you can to the level presented in the USDA above, you get the cladogram presented below. Notice that there are some similar names with different suffixes, like Filicopsida below and above or Lycopodiopsida above and Lycopsida below. Notice that both are classes above, but what are they below? "Lycopsid" appears at a different level of the dendrogram than does "Filicopsida." What does "class" mean in the dendrogram below? It turns out that there is no equivalent to class (or family, or order) in the Tree of Life's cladistic approach. So, even though "Filicopsida" includes the same plants in both schemes, it fits into the overall scheme very differently in each approach to taxonomy.

The conclusion to draw from this is that there are different taxonomic methods to serve different purposes. The standard hierarchy is preferred as an organizational scheme - a good filing system. Using it, one can quickly find plants with particular characteristics. The standardization of the hierarchy (Kingdom, Division, etc.) is familiar and easy to visualize. However, it is not a truly systematic approach. Attempts to make it more natural fail when the actual groups can't be easily fitted into the hierarchical scheme without making loads of new levels (superfamilies, subclasses, sections, groups, etc.) which makes the scheme too cumbersome. Cladistics are the best method we have of presenting the evolutionary history of an organism because of the flexibility of the bifurcating tree. However, the trees are not easy to memorize and it may be difficult to find any phenotypic (morphological, anatomical or chemical) characteristics that separate and uniquely define a group that is formed by analyzing the genotype (the gene sequences). In addition, there are evolutionary events that are difficult to represent with a bifurcating tree (new species arising from hybridization is one). No scheme is perfect at this time and a perfect scheme is unlikely in the future.

A note on the taxonomic levels used in botanical classification. Most should be familiar to you from animal taxonomy: species, genus, family, order, and class. The only difference is the use of "division" by botanists rather than "phylum" for the level below that of kingdom. There is a reason for this change. A phylum represents a basic "body plan" in the animal kingdom. All rotifers have key characteristics that form the rotifer body plan, as all chordates must have the basic chordate body plan. However, related plants do not seem to have a basic body plan. The pea family contains herbaceous members, like garden beans and peas, and members that are trees, like the honey locust trees that bloom so profusely in Tennessee. So, what's the body plan here? Hard to say, so the idea that plants can be separated into basic body plans never took hold and the use of the turn "phylum" would be inappropriate. Instead, classic plant taxonomy relied on the anatomy of the reproductive structures only (that's how locust trees and beans got into the same groups - just look at the pods they each produce).

There is an excellent free resource for learning more about cladistics. "The Compleat Cladist" (the title is a take-off of Izaak Walton's 17th century meditation on fishing, "The Compleat Angler" - hence the odd spelling) by E. O. Wiley, D. Siegel-Causey, D. R. Brooks, and V. A. Funk (1991, Special Publication No. 19, The University of Kansas Museum of Natural History, Lawrence, KS) is both a primer on how to construct trees and a good introduction to the theory behind cladism. Since there is to be a second edition, the authors have been distributing the entire book as a pdf file. It can be obtained at or by emailing me and requesting I send a copy of my copy. As with any pdf file, you need a reader application to view it. The easiest to get is Acrobat Reader from http://www.adobe.com if you computer doesn't already have it. It is freeware.

`Academic``Page`	`Tennessee State` `Home page`
`Bio 432` `Page`	`Ganter``home page`

		Leaf Size
		5 inch	6 inch
Cup Type	rough cup	Species A	Species B
Cup Type	smooth cup	Species C	Species D